To review the pharmacokinetic literature regarding valproic acid (VPA) in patients with bipolar affective disorder to evaluate the appropriateness of routine therapeutic drug monitoring (TDM) of VPA in this population. Embase, Medline, Biosis, Cumulated Index to Nursing and Allied Health Literature (CINAHL), and Cochrane Systematic Reviews databases (to July 2009) were searched (terms: valproate, valproic acid, valproate semisodium, divalproex, bipolar disorder, mania, bipolar affective disorder, assay, concentration, drug level, serum level, plasma level, therapeutic drug monitoring, drug monitoring, pharmacokinetics). A 9-step decision-making algorithm was applied to the available VPA literature in the bipolar population. VPA has been established as effective first-line therapy in the treatment of bipolar mania. Commercial assays are available to accurately measure VPA levels with good sensitivity and precision. Wide interpatient variability in VPA pharmacokinetic parameters and long anticipated treatment durations favor the use of TDM in the bipolar patient. Conversely, correlations between VPA levels and pharmacologic response in humans, and specifically patients with bipolar affective disorder are unclear. Compared with other agents known to have narrow therapeutic ranges, VPA does not seem to share this characteristic. Furthermore, the clinical efficacy and toxicity of VPA may be assessed in ways other than TDM, allowing for contributions by health care professionals, patients, and family members. Despite the development of effective and widely available VPA assays, routine monitoring of VPA concentrations in the general bipolar affective population does not seem warranted. However, TDM may enhance usual clinical monitoring in situations involving unusual VPA metabolism, polytherapy with clinically relevant drug interactions, or as part of a comprehensive assessment of treatment compliance.